ePAD, an oocyte and early embryo-abundant peptidylarginine...

ePAD, an oocyte and early embryo-abundant peptidylargininedeiminase-like protein that localizes to egg cytoplasmic sheets

Paul W. Wright,a Laura C. Bolling,a Meredith E. Calvert,a Olga F. Sarmento,a

Elizabeth V. Berkeley,a Margaret C. Shea,a Zhonglin Hao,a Friederike C. Jayes,a

Leigh Ann Bush,a Jagathpala Shetty,a Amy N. Shore,a Prabhakara P. Reddi,a

Kenneth S. Tung,b Eileen Samy,b Margaretta M. Allietta,a Nicholas E. Sherman,c

John C. Herr,a and Scott A. Coonroda,*a Department of Cell Biology and Center for Research in Contraceptive and Reproductive Health,

University of Virginia, Charlottesville, VA 22908, USAb Department of Pathology, University of Virginia, Charlottesville, VA 22908, USA

c W.M. Keck Biomedical Mass Spectrometry Laboratory, University of Virginia, Charlottesville, VA 22908, USA

Received for publication 22 October 2002, revised 16 December 2002, accepted 17 December 2002

Abstract

Selected for its high relative abundance, a protein spot of MW �75 kDa, pI 5.5 was cored from a Coomassie-stained two-dimensionalgel of proteins from 2850 zona-free metaphase II mouse eggs and analyzed by tandem mass spectrometry (TMS), and novel microsequenceswere identified that indicated a previously uncharacterized egg protein. A 2.4-kb cDNA was then amplified from a mouse ovarianadapter-ligated cDNA library by RACE-PCR, and a unique 2043-bp open reading frame was defined encoding a 681-amino-acid protein.Comparison of the deduced amino acid sequence with the nonredundant database demonstrated that the protein was �40% identical to thecalcium-dependent peptidylarginine deiminase (PAD) enzyme family. Northern blotting, RT-PCR, and in situ hybridization analysesindicated that the protein was abundantly expressed in the ovary, weakly expressed in the testis, and absent from other tissues. Based onthe homology with PADs and its oocyte-abundant expression pattern, the protein was designated ePAD, for egg and embryo-abundantpeptidylarginine deiminase-like protein. Anti-recombinant ePAD monospecific antibodies localized the molecule to the cytoplasm ofoocytes in primordial, primary, secondary, and Graafian follicles in ovarian sections, while no other ovarian cell type was stained. ePADwas also expressed in the immature oocyte, mature egg, and through the blastocyst stage of embryonic development, where expression levelsbegan to decrease. Immunoelectron microscopy localized ePAD to egg cytoplasmic sheets, a unique keratin-containing intermediate filamentstructure found only in mammalian eggs and in early embryos, and known to undergo reorganization at critical stages of development.Previous reports that PAD-mediated deimination of epithelial cell keratin results in cytoskeletal remodeling suggest a possible role for ePADin cytoskeletal reorganization in the egg and early embryo.© 2003 Elsevier Science (USA). All rights reserved.

Introduction

During growth, the oocyte accumulates a pool of mater-nal gene products and organelles required for early devel-opment. In the fully grown egg, the transcriptional machin-

ery is silent and, once ovulated, the terminally differentiatedegg will die if it does not bind and fuse with a sperm. Iffertilization occurs, however, maternal gene products or-chestrate the transformation of the egg into a totipotentzygote within several hours. Following gamete fusion, cal-cium transients propagated in the egg cytoplasm lead to theactivation of signal transduction cascades, which arethought to mediate early embryonic events, such as remod-eling of the cortical cytoskeleton, cortical granule exocyto-sis, completion of meiosis, polar body emission, and for-

* Corresponding author. Department of Cell Biology, Box 439, Uni-versity of Virginia Health System, Charlottesville, VA 22908, USA. Fax:�1-434-982-3912.

E-mail address: [email protected] (S.A. Coonrod).

R

Available online at www.sciencedirect.com

Developmental Biology 256 (2003) 73–88 www.elsevier.com/locate/ydbio

0012-1606/03/$ – see front matter © 2003 Elsevier Science (USA). All rights reserved.doi:10.1016/S0012-1606(02)00126-4

mation of the male and female pronuclei (Yanagimachi,1994). Reprogramming of the parental chromatin is alsothought to occur soon after fertilization, and by the two-cellstage, maternal transcripts begin to be replaced by embry-onic transcripts, and the embryonic genome is activated(Schultz et al., 1999). Maternal factors, however, continueto persist in the early embryo until at least the morula stageof development (Pratt et al., 1983).

Many of the structural and molecular mechanisms me-diating the physiological changes in the early embryo are asyet incompletely characterized. One of the most abundantcytoskeletal components of the mammalian egg is a fibrousnetwork of intermediate filaments (Gallicano et al., 1994a;Lehtonen, 1985, 1987; Lehtonen et al., 1983; Uranga et al.,1995) named the cytoplasmic sheets (Capco et al., 1993).These organelles were previously thought to be either yolkplatelets or possibly ribosome storage sites (Yanagimachi,1994); however, electron microscopic studies indicate thatthe highly ordered sheets are composed of parallel arrays of�10-nm fibers (Gallicano et al., 1991). This filamentousnetwork is stabilized by cross-bridges and is overlain with atightly packed particulate material. Solubility and immuno-logical studies indicate that the Tween 20 insoluble cross-linked fibers contain keratin (but not vimentin or tubulin)and the soluble fraction largely consists of an unidentified�69-kDa protein (Capco et al., 1993). Soluble protein ki-nase C associates with the cytoplasmic sheets, phosphory-lates cytokeratin and the �69-kDa soluble protein, and maybe responsible for initiating the changes in spatial organi-zation that these sheets undergo at the time of fertilization(Gallicano and Capco, 1995). Cytoplasmic sheets arise dur-ing oocyte development (Gallicano et al., 1994b), areunique to the egg and early embryo, are conserved amongmammals (Gallicano et al., 1992), and undergo extensivespatial reorganizations during the critical developmentaltransitions of fertilization, compaction, and blastulation(Capco and McGaughey, 1986).

Peptidylarginine deiminases (PADs) are a family of cal-cium-dependent sulfhydryl enzymes that convert arginineresidues to citrulline in proteins (Senshu, 1990). PAD ac-tivity appears to be upregulated by a variety of estrogeniccompounds (Nagata et al., 1990; Senshu et al., 1989; Taka-hara et al., 1992), and to date, four types of PADs have beencharacterized with each differing in its pattern of substratespecificity and tissue distribution. For example, the widelydistributed and well characterized type II PAD is especiallyabundant in muscle (Takahara et al., 1983, 1986) and brain(Akiyama et al., 1999; Asaga and Ishigami, 2001) and isassociated with deimination of myelin basic protein (La-mensa and Moscarello, 1993; Moscarello et al., 2002). PADV, found in granulocyte-differentiated HL-60 cells, isthought to play a role in myeloid cell differentiation (Na-kashima et al., 1999), and likely targets nucleophosmin andhistone core proteins for deimination (Hagiwara et al.,2002). Type I and III PADs have been characterized in theepidermis; with type III PAD being found to deiminate

trichohyalin in hair follicles (Kanno et al., 2000; Nishijyo etal., 1997) and Type I PAD deiminating keratin and filaggrinduring epidermal differentiation (Akiyama and Senshu,1999; Senshu et al., 1999a). It is thought that the deimina-tion of keratin and filaggrin in the epidermis induceschanges in the spatial organization of keratin intermediatefilaments during keratinocyte maturation (Ishida-Yamamotoet al., 2002). Deiminated keratin has been identified in day18 embryos (Akiyama and Senshu, 1999); however, thepresence of deiminated keratin at earlier stages of develop-ment has not been investigated.

In an ongoing project to identify previously uncharacter-ized proteins in the ovulated mouse egg, we have beenperforming tandem mass spectroscopic analysis of abundantoocyte proteins that have been cored from Coomassie-stained two-dimensional (2D) gels. One dominant egg pro-tein spot that yielded novel peptide microsequences wascloned and the cDNA characterized. Using Northern blot-ting, the tissue distribution was determined and the proteinwas found to be expressed in primary oocytes and to persistuntil at least the blastocyst stage of development. Due to its40% identity with the peptidylarginine deiminase enzymefamily, the name ePAD, for egg and embryo abundant PAD,was selected. Remarkably, at the ultrastructural level, ePADlocalized to the egg’s cytoplasmic sheets, a cytoskeletalstructure unique to the egg and early embryo. The discoverythat ePAD is associated with the egg cytoplasmic sheetsleads to the hypothesis that arginine deiminase reactionsdirected against cytokeratin, and possibly other proteins,results in reorganization of the cytoskeleton during earlydevelopment.

Materials and methods

Two-dimensional electrophoresis

Mouse oocytes (2850 for the Coomassie-stained gel inFig. 1 and �300 per blot in Fig. 6) were collected anddezonulated as described previously (Coonrod et al., 1999).The zona-free eggs were then washed six times in PBScontaining 10 �g/ml polyvinylalcohol (PVA; Sigma) andextracted in Celis lysis buffer [containing 2% (v:v) NP-40,9.8 M urea, 100 mM dithiothreitol (DTT), 2% ampholines(pH 3.5–10), and protease inhibitors] for 30 min at roomtemperature (Rasmussen et al., 1991). Isoelectric focusing(IEF) was performed by using the BioRad Protean II Multi-Cell apparatus with an ampholine mixture (Pharmacia Bio-tech, Uppsala, Sweden) of pH 3.5–5 (30%), 3.5–10 (40%),5–7 (20%), and 7–9 (10%). The tube gels were placed on12% slab gels (16 � 16 cm, plates 1.5 mm diameter), andthe focused proteins were separated in the second dimen-sion. The gels were then either stained with Coomassie orelectroblotted to nitrocellulose membranes. For Coomassiestaining, the gels were fixed overnight in a solution of 50%ethanol and 10% acetic acid and placed in a solution con-

74 P.W. Wright et al. / Developmental Biology 256 (2003) 73–88

taining 0.1% Coomassie R250, 40% methanol, and 0.1%acetic acid for 4 h. The gels were then destained in asolution of 10% acetic acid and 50% methanol.

Tandem mass spectroscopic analysis of egg peptides

Two �75-kDa (pI 5.5) Coomassie-stained protein spotswere cored from the 2D SDS-PAGE gel, fragmented,destained in methanol, reduced in 10 mM dithiothreitol, andalkylated in 50 mM iodoacetamide in 0.1 M ammoniumbicarbonate. The gel pieces were then incubated with 12.5ng/ml trypsin in 50 mM ammonium bicarbonate overnightat 37°C. Peptides were extracted from the gel pieces in 50%acetonitrile and 5% formic acid and microsequenced bytandem mass spectrometry at the Biomolecular ResearchFacility of the University of Virginia.

Adapter-ligated cDNA library construction and RACE-PCR cloning of ePAD cDNA

Two micrograms of mouse ovarian poly(A)� mRNA,isolated using the FastTrack 2.0 kit from Invitrogen (Carls-bad, CA), was used as the template for the construction ofa Marathon adaptor ligated cDNA library (Clontech, PaloAlto, CA). Oligo(dT) primers, as well as avian myeloblas-tosis virus (AMV) reverse transcriptase, were used to con-struct the first strand of cDNA. The RNA was digested andthe second strand of cDNA was synthesized in a simulta-neous reaction with a mixture of Escherichia coli DNApolymerase I, Rnase H, and E. coli DNA ligase. The cDNAends were then blunted by using T4 DNA polymerase, andMarathon cDNA adaptors were ligated to both ends of thecDNA by the addition of T4 DNA ligase.

RACE PCRs were performed using the Amplitaq GoldDNA polymerase from Perkin-Elmer (Norwalk, CT) withovarian library cDNA templates. Cycling parameters were:94°C, 10 min; 94°C, 15 s; 60°C, 30 s; 72°C, 2 min; and72°C, 10 min, for 40 cycles. PCR of plasmid templates wasperformed under the same conditions, except that PromegaAdvantage 2 taq polymerase was used. The primers forcloning the full-length ORF of ePAD into the TOPO clon-ing vector (Invitrogen) were 5�-GTACTCGAGATGGTAG-GCATGGAAATCACC-3� and 5�-CGATCTAGAT-GGGGTCATCTTCCACCACTT-3�, and the AP1 primersupplied in the Marathon ready cDNA kit (Clontech).

Northern blot analysis

A randomly primed probe was generated correspondingto the 1100-bp N-terminal region of ePAD by using thePrime-a-Gene Labeling System (Promega, Madison, WI).Twenty-five nanograms of cDNA were denatured at 95°Cfor 5 min. After cooling on ice for 5 min, 50 mM Tris–HCl,pH 8.0, 5 mM MgCl2, 2 mM DTT, 200 mM Hepes, pH 6.6,random hexadeoxyribonucleotides, 400 �g/ml of BSA, 333nM [�-32P]dCTP, 20 �M each of dATP, dGTP, and dTTP,

and 5 units of DNA polymerase I, large (Klenow) fragment,were added to the cDNA. This was incubated for 2 h at22°C, and subsequently passed through a NucTrap (Strat-agene, La Jolla, CA) column to remove unincorporatednucleotides from the probe. A mouse multi-tissue Northernblot (Clontech) was incubated with ExpressHyb hybridiza-tion solution (Clontech) for 1 h at 65°C. The probe washeated to 95°C for 5 min, and placed in 10 ml of freshExpressHyb. This was incubated with the blot overnight at55°C, and washed on the following day, twice for 15 minwith 2� SSC and 0.5% SDS, twice for 30 min in 0.1� SSCand 0.5% SDS at 65°C. Signals were detected by exposureto X-ray film for 24 h and for 10 days.

RT-PCR of ePAD in mouse tissues

The Mouse Multiple Tissue cDNA Panel from Clontech,comprised of cDNA from 11 different tissues, served astemplate in the PCRs. cDNAs from the Panel were used toamplify ePAD and control G3PDH cDNA fragments in50–�l reactions (10 mM Tris–HCl, 1.5 mM MgCl2, 50 mMKCl, 0.1% Triton X-100, 0.2 mM of each dNTP, 0.5 U ofTaq polymerase, and 0.5 mM of each primer). Thermocy-cler settings consisted of a 10-min incubation at 95°C,followed by 40 cycles at 95°C (15 s), 60°C (30 s), and 72°C(2 min). Extension progressed for 10 min at 72°C. Tenpercent of product was later analyzed on 1.5% agarose gelsand confirmed by Southern blotting. The following ePADprimers were used: 1–27, 5�-GGTAAGGACTGCTGA-CAGTGGC-TAGCT-3�; 2222–2249, 5�-CCTCCAGAAG-TGTCAAGCTGGTAGATGT-3�. G3PDH primers were in-cluded in the previously mentioned Clontech kit.

In situ hybridization

Ovaries from fertile ICR female mice were fixed inneutral buffered formalin solution (4%) (Sigma, St. Louis,MO) for 12 h. Following dehydration, the blocks wereembedded in paraffin, sectioned (2.5 mm), and mounted onpositively charged slides. To prepare a riboprobe, a DNAfragment consisting of the 5� 250 nucleotides of the ePADORF was subcloned into the pBluescript SK plasmid vectorand used as template for in vitro transcription (Angerer etal., 1987). Tritiated Q2 uridine triphosphate (UTP) wasincorporated into the riboprobes by either T3 (sense) or T7(antisense) RNA polymerase. A labeled �-actin riboprobewas used as a positive control. The sections were deparaf-finized, rehydrated, and treated with proteinase K. The insitu hybridization solution contained 50% formamide, 0.3M NaCl, 20 mM Tris–HCl, 1 mM EDTA (pH 8.0), Den-hardt’s solution, 500 mg/ml yeast tRNA, and 10% dextransulfate. The final probe concentration was normalized forprobe length and applied at full saturation (0.2 mg/ml/kbcomplexity). The hybridization was carried out at 55–65°C.After hybridization, the sections were washed under high-stringency conditions to remove nonspecific hybridization.

75P.W. Wright et al. / Developmental Biology 256 (2003) 73–88

The slides were overlaid with autoradiograhy emulsion,exposed 2–4 weeks at 48°C, developed photographically,and lightly stained with hematoxylin and eosin. Dark fieldimages were then captured at 200� by using a Zeiss mi-croscope.

Expression of recombinant protein and antibodyproduction

Primers were designed to generate a PCR product en-coding amino acids 1–200 of the open reading frame. Thesense primer, 5�-TAAGGATCCGATGGTAGGCATGGA-AATCACCTTGG-3�, contained an engineered BamHI re-striction site, while the anti-sense primer, 5�-GTACTC-GAGCCAGCCTCTCTGAGACCAGTAGACTCT-3�, hada XhoI restriction site. The primers were used to amplify thecDNA fragment, and the PCR product was cloned into theBamHI–XhoI sites of the pET22b expression vector (Nova-gen, Madison, WI). Once the construct was verified, it wastransformed into E. coli strain BL21 DE3 cells (Stratagene).A single positive colony was used to inoculate a 10-Lculture that was grown at 37°C in Luria broth (LB), in thepresence of 50 mg/mL ampicillin until the A600 reached0.6. Recombinant protein expression was induced with 1.0mM isopropyl-1-thio-�-D-galactopyranoside (IPTG) for 3 h.

Protein was purified from the cells by Ni-affinity chroma-tography, which binds to a 6� His tag attached to therecombinant protein (Reddi et al., 1994). Half of the bio-mass was suspended in 5 mM imidazole, 0.5 M NaCl, and20 mM Tris–HCl, pH 7.9, and was sonicated four times for45 s. The sonicated extract was centrifuged at 20,000g for15 min to collect inclusion bodies and cellular debris. Thesupernatant was removed, and the pellet was then resus-pended in 6 M urea, 5 mM imidazole, 0.5 M NaCl, and 20mM Tris–HCl, pH 7.9. This was incubated on ice for 1 h tocompletely dissolve the protein. The remaining insolublematerial was removed by centrifugation at 39,000g for 20min and filtration through a 0.45-micron membrane. Theremaining supernatant was bound overnight at 4°C to nickelcharged by His-Bind Resin (Novagen) while shakinggently. After centrifugation, the resin was washed twice for20 min with 5 mM imidazole, 0.5 M NaCl, and 20 mMTris–HCl, pH 7.9, and once with 20 mM imidazole, 0.5 MNaCl, and 20 mM Tris–HCl, pH 7.9, for 20 min. Theproduct was eluted with 1 M imidazole, 0.5 M NaCl, and 20mM Tris–HCl, pH 7.9, for 20 min and collected. Elute fromthe column was passed through a PREP cell (BioRad), inorder to remove remaining contaminants. Final product wasthen dialyzed against 1.0 mM ammonium bicarbinate for36 h before lyophilization and reconstitution.

Polyclonal antibody production and IgG purification

Three adult male Hartley guinea pigs (albino retiredbreeders) were used for antibody production against a par-tial fragment (extreme 200 N-terminal amino acids) en-coded in the pET22b/ePAD clone. Preimmune serum wascollected by heart puncture, and subsequently, each animalwas injected with 100 �g of purified recombinant PAD(rePAD) in an emulsion with Freund’s complete adjuvant.Each animal received two booster immunizations with 50�g of purified rePAD in Freund’s incomplete adjuvant at3-week intervals. For all immunizations, half of the antigenemulsion was injected intramuscularly in the leg and halfsubcutaneously in three sites on the back. All animals werethen exsanginated by heart puncture 10 days after the finalimmunization, and the blood was collected in serum sepa-ration tubes (Becton Dickinson, Franklin Lakes, NJ). Aftercentrifugation at 1750g for 10 min, the serum was removedand frozen until needed. A prepacked Protein A columnfrom Pierce was used for purifying IgGs from anti-rePADmouse serum. The column was prewashed with 5 mL ofsterile water. To calibrate the column, 5 mL of bindingbuffer supplied by Pierce was then added and allowed todrain. Sera was diluted 1:10 with binding buffer, passedthough the column in 4-mL increments, and the bound resinwas washed with binding buffer. IgGs were eluted from thecolumn with elution buffer from Pierce and dialyzed againstPBS before quantitating.

Fig. 1. Location of ePAD on a Coomassie-stained 2D electrophoretic gel of2850 NP-40 extracted zona-free mouse oocytes. Arrows indicate the twoprotein spots that were cored from the gel for TMS analysis. The putativeePAD protein chain is delineated by the brackets.


Anti-ePAD antibody specificity and ePAD solubilitycharacteristics

BL21 cells expressing rePAD were sonicated and 20 �gbacterial protein, 50 oocytes, and one million sperm per lane

were solubilized in Laemmli buffer (Laemmli, 1970) andheat denatured at 95°C. The samples were then loaded ontoa 12.5% linear gel, and separated at 100 V for 3 h. Proteinsthen electrotransferred onto 0.2 �m nitrocellulose (Bio-RadLaboratories, Hercules, CA) for 40 min at 100 V for West-ern blotting.

To investigate solubility characteristics, 70 mouse eggswere extracted with NP-40 lysis buffer (150 mM sodiumchloride, 1.0% NP-40, and 50 mM Tris, pH 8.0), incubatedon ice for 30 min, and passed through a 25-G needle threetimes. The extract was then centrifuged for ten minutes at10,000g. The supernatant and pellet were each solubilizedin Laemmlie buffer, heat denatured at 95°C, and separatedon a 12.5% SDS-PAGE gel under reducing and nonreduc-ing (lacking �-mercaptoethanol) conditions. Protein wasthen transferred onto a nitrocellulose and blotted as de-scribed below.

Western blot analysis

For Western blot analysis of ovarian ePAD expression,ovaries from 2, 4, 8, 15 wk, and retired breeder femaleswere homogenized in extraction buffer (PBS, 1% NP-40,and 1� concentration of Complete, EDTA-free ProteaseInhibitor Cocktail Table; Roche, Mannheim, Germany). Af-ter homogenization, the samples were sonicated for 30 seach and quantified using the Biorad DC Protein Assay kit(Biorad). Equal amounts of each sample (21.5 �g) wereloaded on a 12.5% SDS-PAGE gel, and separated at 115 V.Protein was then blotted to nitrocellulose, and stained withPonceu.

All blots were blocked with 5% nonfat dry milk in PBSwith 0.05% Tween 20 (PBS-T) for 30 min, washed andincubated with a 1:5000 dilution of anti-rePAD guinea pigsera. The blots were then washed two times for 10 min inPBS-T, and incubated with a 1:10,000 dilution of peroxi-dase conjugated goat anti-guinea pig IgG (Jackson Immu-noResearch, West Grove, PA) secondary antibody for 1 hand washed two times for 10 min in PBS-T. The blots werethen either developed in TMB peroxidase substrate (3,3�,5,5�-tetramethylbenzidine; Kirkegaard and Perry Laborato-ries, Gaithersburg, MD) or with ECL reagent (AmershamCorp., Buckinghamshire, UK) and developed as describedpreviously (Coonrod et al., 1999).

Indirect immunofluorescence of ovarian sections, ovaries,and embryos

Cryosections (5 �m) of mouse ovary were processed forindirect immunofluorescence. Slides were immersed in 95%ethanol for 10 min, washed, and incubated in a blockingbuffer of PBS which contained 3% BSA and a 1:50 dilutionof normal goat serum for 20 min at room temperature. Next,the slides were incubated with blocking buffer which con-tained either a 40-�g/ml ePAD IgG or preimmune IgG for1 h at room temperature, washed, incubated in blocking

Fig. 2. Complementary DNA and deduced amino acid sequence of ePAD.The untranslated flanking regions are shown in lower case letters. Peptidesequences originally obtained by mass spectrometry are underlined.


buffer containing goat anti-guinea pig IgG-FITC (1:200dilution), washed again, and the slides were coverslippedand stored at 4°C prior to imaging. All oocytes and embryos

were obtained from ICR 25–30 g females (Hilltop Labs).Germinal vesicle oocytes and metaphase II eggs were ob-tained as previously described (Coonrod et al., 1999, 2001).

Fig. 3. Amino acid sequence comparison of ePAD with the four known mouse peptidylarginine deiminases (PADs). The sequences were aligned by usingthe ClustalW alignment program in BioEdit sequence alignment editor. Identical residues are shown in black boxes, and conserved residues are shown ingray boxes. Dashes indicate gaps inserted to maximize sequence alignment. PAD Accession Nos. are as follows: type I, NP_035198; type II, NP_032838;type III, NP_035190 type IV, NP_035191; ePAD, NP_694746. PADs I–IV are approximately 50% identical to each other, while ePAD is approximately 40%identical to known murine PADs.


Pronuclear embryos were obtained from females after in-jection of 10 I.U. of PMSG followed 48 h later by 10 I.U.of hcG and mating with ICR male retired breeders. Eighteento twenty-four hours after injection of hCG, pronuclearstaged embryos were isolated from the oviduct in Whitten’smedia with Hepes (Specialty media; Phillipsburg, NJ) con-taining 0.05% hyaluronidase. Embryos were washed thor-oughly in TE media and either removed for fixation orallowed to develop in vitro at 37°C and 5% CO2. Oocytesand embryos were fixed in 4% paraformaldehye in PBS for20 min at room temperature. Following fixation, oocytesand embryos were washed five times in PBS � 1% BSA(PBS/BSA) and then permeabilized with 0.5% Triton X-100in PBS. Oocytes and embryos were then washed five timesin PBS/BSA and blocked for 30 min with PBS/BSA con-taining 10% goat serum. Oocytes and embryos were thenincubated with 40 �g/ml ePAD IgG or 40 �g/ml guinea pigIgG for 1 h at room temperature. Oocytes and embryos werewashed five times and incubated for 1 h at room temperaturewith goat anti-guinea pig FITC-labeled secondary antibody(Jackson ImmunoResearch). Oocytes/embryos were mountedon slides and visualized under a Zeiss Standard 18 ultravioletmicroscope.

Confocal microscopy

Eggs and embryos used for confocal microscopy wereprepared as above except that in vivo derived embryos wereused. The morula and blastocyst staged embryos were ob-tained by flushing the oviducts of mated females either 4 or5 days following hCG injection and processed as above.Microscopy images were obtained on a Ziess Axiovert 100micro systems LSM confocal microscope. Each image wasacquired using 4 s scan averaged four times per line using a40� oil lens with a zoom of approximately 2. False colorwas added as appropriate.

Electron microscopy

Mouse ovaries were fixed for 2 h at room temperature in8% paraformaldehyde buffered with 0.1 M sodium phos-phate, pH 7.5. The ovaries were processed through a seriesof ethanol washes to dehydration and embedded with Lowi-cryl resin (Electron Microscopy Sciences, Fort Washington,PA). Ultrathin sections were picked up on 75 mesh nickelgrids and immunolabeled with primary antibody overnightat 4°C, washed, and secondary antibody on 5 nm colloidalgold was added to sections at room temperature for 45 min,washed, and heavy metal-stained with aqueous saturatedUranyl acetate for 30 min. The sections were viewed andphotographed by using a JEOL 100CX2 electron micro-scope.

Results

Tandem mass spectroscopic identification of ePAD, ahighly abundant novel egg protein

Zona-free mouse eggs (2850) were extracted, separatedon a 2D electrophoretic gel, and stained with Coomassie(Fig. 1). Within a train of proteins (pI � 5 to 5.5) two spots,designated ePAD (Fig. 1, arrows) and migrating at �75kDa, were cored from the gel, digested with trypsin, andmicrosequenced by CAD Mass spectrometry. Twenty-eightpeptides were identified in the spot indicated by the arrowon the left in Fig. 1, and 12 peptides were identified in thespot indicated by the arrow on the right, with 5 commonpeptides being shared between the 2 spots. The presence ofcommon peptides within the 2 spots suggested that theentire protein train (Fig. 1, brackets) may arise from a singlegene or related genes. Comparison of the staining intensityof this protein train with that of other protein spots foundthat ePAD appears to be one of the most abundant proteinsin the ovulated mouse egg. When the identified peptideswere searched against nonredundant database sequences, itwas concluded that this protein had not been previouslycharacterized. One of the common peptides having the se-quence CCLEEKVCGLLEPLGLK, however, matched 2EST clones arising from 2-cell (AA645498) and 8-cell(AU020751) mouse embryo cDNA libraries.

Cloning and characterization of ePAD

Repeated RACE PCR was performed by using an ovar-ian Marathon-ready adapter-primed cDNA library and oli-gonucleotide primers designed from the CCLEEKVCG-LLEPLGLK peptide sequence, the matching EST clones, orgene-specific sequences resulting from amplification of 5�and 3� PCR gene fragments. Oligonucleotides were de-signed from the gene fragments and used to PCR-amplify a2374-bp cDNA from the ovarian library. The gene wasinserted into a cloning vector and subsequent DNA se-quencing revealed a contiguous open reading frame of 2043bp which encoded a 681-amino-acid protein (Fig. 2). Theuntranslated flanking regions are shown in lower case let-ters. The GenBank Accession No. for this clone isNP_694746. The computed mass and isoelectric point of thededuced amino acid sequence is 76.7 and 5.36, respectively(ExPASy compute pi/MW tool), which closely matched theapproximate mass and pI of the protein spots that werecored from the gel for TMS analysis. Nineteen of the 28peptides originally identified by TMS were found in thededuced amino acid sequence and are underlined. When thededuced ePAD amino acid sequence was searched againstthe Prosite computational database, no highly significantstructural motifs were observed. An NCBI conserved do-main search using the deduced amino acid sequence, how-ever, identified a peptidylarginine deiminase (PAD) domain(E value � 0.0) that spanned the protein sequence from the


seventh residue to the C terminus. Comparison of the se-quence against the NCBI nonredundant database using theBLAST algorithm found that the sequence was 96% iden-tical to a recently submitted putative mouse PAD typeV-like protein sequence (XM_144067), which was pre-dicted by NCBI’s automated sequence analysis program.The next most similar sequence was a predicted humanputative human PAD type V-like protein sequence(XP_017315.3) with an identity of 63%. ePAD was �40%identical to all other characterized PAD sequences, includ-ing those of mice, humans, rats, sheep, and chickens. Adirect comparison of ePAD with the four characterizedmouse PADs using the ClustalW alignment function in theBioEdit sequence alignment editor indicated that ePAD isapproximately 40% identical to the known PADs, whileeach of these PADs is approximately 50% identical to eachother (Fig. 3). Interestingly, most of the domains that areconserved between the known PADs are also conserved inePAD, indicating that ePAD likely represents a new mem-ber of the PAD enzyme family.

ePAD is expressed in oocytes and early cleavage-stageembryos

An EST database search using the ePAD open readingframe nucleotide sequence found nearly identical matchesin mouse ovary, unfertilized egg, fertilized egg, 2-cell em-bryo, 8-cell embryo, and more recently, the day 0 neonatalthymus (data not shown). This information led to the hy-pothesis that ePAD may be an egg-abundant transcriptwhose expression persists throughout the early stages ofcleavage. ePAD expression was examined in various tissuesby using Northern blotting, RT-PCR, and in situ hybridiza-

tion. A multitissue Northern blot (Clontech) containing 5�g poly(A)� RNA, and a separate blot containing 40 �g oftotal ovarian RNA, were probed with a 32P-labeled random-primed 1100 bp ePAD N-terminal DNA fragment. A single2.4-kb message was found to be abundantly expressed in theovary (Fig. 4A). Subsequent multiple tissue Northern blot-ting experiments using a single membrane showed similarresults (data not shown). Ovary-abundant expression ofePAD was further confirmed by RT-PCR and Southernblotting (Fig. 4B). A mouse multiple tissue cDNA panel(Clontech) comprised of cDNA from 11 different tissues(including day 7 and day 15 embryos) was used as templatein the PCRs. The oligonucleotide primers used for the PCRsencompassed the ePAD ORF (2043 bp). A strong �2-kbamplicon was detected in the ovary cDNA lane, a weakeramplicon was seen in the testis cDNA lane, and no ampli-cons were detected in lanes containing cDNA from othertissues. To confirm the authenticity of the PCR ampliconsseen in the ovary and testis lanes, the PCR products wereblotted to a nitrocellulose membrane and the blot wasprobed with the same 32P-labeled ePAD probe, which wasused for Northern blotting. Results showed that the PCRproducts were derived from ePAD sequence (Fig 4B.). A32P-labeled full-length �-actin probe and olignulceotidesdesigned to amplify the open reading frame of G3PDH wereused as positive controls for the Northern blot and theRT-PCR, respectively.

In situ hybridization experiments were then performed tolocalize ePAD expression within the ovary. Results showedePAD message only in oocytes (Fig. 4C) and not in otherovarian cell types. Further, ePAD message was not detectedin oviductal or testis cross sections (data not shown). Takentogether, these results suggest that ePAD is abundantlyexpressed in the ovary (specifically within oocytes) and maybe expressed at lower levels in testis and thymus.

Fig. 4. Tissue distribution of ePAD. Northern blot analysis was performed on various tissues by using a P32-labeled 1100-bp ePAD DNA fragment (A). TheNorthern membrane contained 40 �g of total RNA for the ovary lane and 5 �g poly (A) RNA for the other tissues (exposure time 24 h). ePAD was detectedas a single �2.4-kb band in the ovarian RNA lane. A �-actin probe was used to control for RNA loading. RT-PCR and Southern blot analysis of ePAD tissuedistribution (B). Specific primers based on the ePAD open reading frame were used in PCRs to screen cDNA from various tissues for ePAD expression.cDNA was tested from the following tissues: heart (Ht.), brain (Br), spleen (Sp), lung (Lu), Liver (Li), skeletal muscle (Sk), Kidney (Ki), testis (Te), day7 embryo (d7), day 15 embryo (d15), and ovary (Ov). Glyceraldehyde 3-phosphate dehydrogenase (G3PDH) primers were used as a positive control. A strongePAD amplicon was seen in the ovarian RNA lane, while a weak amplicon was seen in the testis lane. A Southern blot was performed by using a 32P-labeledePAD DNA fragment to confirm the authenticity of the amplicon. In situ hybridization of mouse ovarian tissue with ePAD anti-sense RNA (C) demonstratedhybridizing to oocytes (white grains against dark field, indicated by arrows) but not to other ovarian cell types, while ePAD sense-strand RNA did notspecifically hybridize to any ovarian tissues. Testis and oviductal sections were also tested with the ePAD probes and no signal was detected (data not shown).Fig. 5. A 75-kDa (pI 5.5) protein spot containing ePAD peptide sequence is specifically recognized by anti-ePAD antibodies. Antisera was generated in maleguinea pigs against a purified recombinant form of ePAD containing the N-terminal 200 amino acids. The preimmune sera was not reactive with either therecombinant protein or with egg (50 mature zona-intact eggs per lane) or sperm proteins (�1 � 106 per lane), while the anti-ePAD sera was reactive withrePAD and with a �75-kDa egg protein, but not with sperm protein (A). The solubility characteristics of ePAD were investigated, and the protein was foundto partition with the 1% NP-10 soluble fraction (So.) and not with the insoluble fraction (In) under both reducing and nonreducing conditions (B). Threehundred ovulated zona-intact mouse oocytes were extracted, separated by 2D electrophoresis, blotted to nitrocellulose, and egg proteins were resolved withprotogold. The blot was then probed with anti-ePAD preimmune IgG and prepared for protein detection by using enhanced chemiluminescence (ECL). Theblot was then washed and reprobed with anti-ePAD immune IgG. Arrows indicate reactive proteins at �75 and 110 kDa. Overlay image localizes reactiveproteins with protein spots resolved using protogold (C). This Western blot demonstrates that the anti-ePAD IgG strongly reacts with the protein train thatwas originally cored for microsequence analysis. In a separate experiment, ePAD peptides were also found by TMS at 110 kDa (data not shown), which mayexplain the observed upper reactive band (upper arrow).


A 75-kDa (pI 5.5) protein spot which contains ePADpeptide sequence is specifically recognized by anti-ePADantibodies generated against N-terminally truncatedrecombinant ePAD

A cDNA encoding an N-terminal 200-amino-acid regionof ePAD was cloned into a bacterial expression vector,expressed, purified, and used as an immunogen for anti-ePAD antisera production in guinea pigs. The reactivity ofanti-ePAD antisera with the recombinant protein and withthe appropriately sized egg protein was first confirmed by1D Western blotting. The preimmune sera was not reactivewith rePAD, egg, or sperm proteins. The immune sera wasreactive with rePAD and an appropriately sized egg protein;however, the antibody did not react with sperm proteins(Fig. 5A). Previous experiments had shown that ePAD wassoluble in several detergents (data not shown); this wasconfirmed by probing Western blots of egg proteins ex-tracted 1% NP-40 and separated under both reducing andnonreducing conditions. As seen in Fig. 5B, ePAD is solu-ble under both conditions. It is of interest that a higher MWePAD immunoreactive region appears under nonreducingconditions, possibly indicating a protein–protein interaction.Given that the anti-ePAD antisera appeared to be of highquality, the IgG fraction of the preimmune and immune serawas purified and used for all subsequent experiments.

On 2D immunoblots, the immune IgG to recombinantePAD reacted with an �75-kDa protein train at the preciselocation from which the two spots were originally cored,while the preimmune IgG was not reactive with any eggproteins (Fig. 5C). This finding provided an immunologicalproof that specific antibodies had been generated against theoriginal cored protein, and that the cDNA and amino acidsequences obtained represented the intended spot. The trainof ePAD isoelectric variants at 75 kDa is likely indicative ofmultiple posttranslational protein modifications. Lesser im-munoreactivity was also noted with a protein train at �125kDa and pI 5.5 (Fig. 5C, upper arrow). In order to determinewhether this immunoreactivity at 125 kDa represented au-thentic ePAD, bands in the 125-kDa MW range were mi-crosequenced from a 1D gel, and ePAD peptide sequences

were noted, confirming that the band at 125 kDa in Fig. 5Crepresents a higher molecular weight form of ePAD (datanot shown).

Ontogenic expression and subcellular localization ofePAD

The monospecific IgG to rec ePAD was employed toinvestigate the developmental expression of ePAD. Ovarieswere collected from 2, 4, 8, and 15 wk females, as well asfrom retired breeders and either embedded for sectioningand indirect immunofluorescence or proteins were extractedfor Western blotting with equal concentrations of ovarianprotein extract loaded in each lane. By Western blot (Fig.6A), the ePAD protein was found at all stages examined.ePAD’s relative abundance, however, appeared to be high-est at 2 weeks of development (the earliest stage tested) andto gradually decrease as the mice aged. Since oocyte num-bers are reduced over time through ovulation and atresia,and since ovarian structures such as corpora lutea and in-terstitial glands differentiate and regress, the ratio of oocytesto total ovarian tissue mass might be expected to vary atsome stages. To control for this possibility, the blots werestripped and reprobed with antibodies to ZP3. Resultsshowed that, as opposed to ePAD, the expression levels ofthis oocyte-specific protein appeared to remain constant orpossibly increase as the ovaries developed and aged. There-fore, the observed decreases at 15 weeks and in retiredbreeders are unlikely to be the result of unequal numbers ofoocytes represented in each sample.

Indirect immunofluorescence of the ovarian cryostat sec-tions confirmed that ePAD was expressed in an oocyte-specific manner in all types of follicles and in all develop-mental stages tested (Fig 6B, stages other than 2 week andretired breeders are not shown). In 2-week-old females,abundant oocyte staining was observed in many primordialfollicles (Pm.) and in all primary follicles (Pr.). When ovar-ian sections of mature females were stained with the anti-ePAD IgG, intense staining was observed in oocytes withinprimary, secondary (Sec.), and some primordial follicles.Staining was also observed in tertiary (Graafian) follicles

Fig. 6. Ontogenic expression of ePAD. Ovaries were collected from mice at different stages of development and either extracted for Western blot analysis(A) or prepared for indirect immunofluorescence (B). (A) Prior to Western blot analysis, blots were probed with Ponceu to control for protein loading. Whenblots of ovarian extracts were probed with anti-ePAD IgG, ePAD expression appeared highest in 2-week-old females and then declined slightly during animalgrowth. The blots were then stripped and reprobed with anti-ZP3 IgG in order to compare ePAD expression with that of another oocyte-specific protein. Asopposed to ePAD, ZP3 expression appeared to increase slightly during animal growth. (B) Ovarian cross-sections from 2-week-old females and maturefemales were probed with anti-ePAD IgG and prepared for indirect immunofluorescence imaging (IIF) and phase contrast microscopy. ePAD expression in2-week-old females was observed in eggs contained within primordial (Pm.) and primary follicles (Pr.). In mature ovaries, ePAD expression is seen in primaryand secondary (Sec.) oocytes, while other ovarian tissues were not stained. Oocytes contained within tertiary follicles were also recognized by anti-ePADIgG (Data not shown). These results further confirm the oocyte-abundant expression pattern of ePAD.Fig. 7. Developmental expression of ePAD in the oocyte and early embryo. Indirect immunofluorescence (IIF) and Western blot (WB) analysis wereperformed to determine whether ePAD continues to be expressed in the developing embryo. Immature (GV), mature (MII), and pronuclear (PN)oocytes/embryos were collected from females, while 2-cell, 4-cell, morula (Mo.), and blastocyst (Bl.) were collected following culture from the pronuclearstage. Indirect immunofluorescence was the performed on the eggs/embryos by using either ePAD preimmune IgG (PI, negative control, MII eggs) or immuneIgG. Western blotting (WB) was also performed on each of the stages shown using 20 oocytes/embryos per group. ePAD appears to be expressed at all stagesshown, with a decrease in expression levels being observed at the blastocyst stage.


(data not shown). No ovarian cell types other than theoocytes were reactive with the antibody.

The expression of ePAD protein during oocyte matura-tion, fertilization, and early embryonic development wasnext investigated. Immature germinal vesicle-stage oocytes(GV), mature metaphase II arrested eggs (MII), and pro-nuclear zygotes (PN) were collected from females and im-mediately fixed. Pronuclear-stage embryos were also col-lected and cultured in vitro, and the cultured embryos wereharvested at the 2-cell, 4-cell, morula (Mo.), and blastocyst(Bl.) stage and fixed for immunocytochemistry. All stageswere then permeablized, stained with secondary antibodyalone, anti-ePAD preimmune IgG, or anti ePAD immuneIgG, and analyzed by indirect immunofluorescence. Nostaining was observed when eggs were stained with anti-ePAD preimmune IgG (Fig. 7, PI). Further, neither theanti-ePAD preimmune IgG or the secondary antibody alonestained any of the other stages tested (data not shown).Anti-ePAD immune IgG immunofluorescence staining in-dicated that ePAD protein was present at all stages, with amore heterogeneous staining pattern seen among the blas-tomeres at the morula stage and an apparent decrease instaining intensity seen at the blastocyst stage (Fig. 7). Totest whether the ePAD protein detected by microscopicmethods could be confirmed biochemically, 20 eggs/em-bryos were collected for each stage as described, extracted,and blotted to a nitrocellulose membrane. The blot, probedwith anti-ePAD antibody, confirmed that ePAD was present

until the blastocyst stage, at which point ePAD expressiondecreased (Fig. 7, WB).

Confocal analysis was then performed to further inves-tigate the subcellular localization of ePAD (Fig. 8). Evalu-ation of immature oocyte (GV) cross sections showed thatePAD localized predominantly to the cortical region of theoocyte, with decreased staining in interior cytoplasmic re-gions. Low levels of ePAD staining were noted in theoocyte nucleus, with complete exclusion of staining fromthe nucleolus. Close examination of the oocyte cytoplasmrevealed that ePAD staining was punctate in nature. Thispattern of punctate staining suggested that ePAD might beassociated with a specific cytoplasmic element. In fertilizedeggs (Fert.), ePAD was expressed in punctate patchesthroughout the cytoplasm. In the morula and blastocyst(Blast.), ePAD appeared to localize mainly to the cyto-plasm; however, a limited amount of nuclear staining wasnoted in some blastomeres. Note that the blastomere indi-cated by the arrow (Fig. 8) appeared to be in metaphase andthat ePAD was excluded from chromatin-containing re-gions. Intense cortical immunofluorescent staining was seenat each of the developmental stages shown, except in themorula, in which cortical staining was limited to the exter-nal surface of the blastomeres and appeared polarized. In theblastocyst, the cortical staining was mainly limited to thecells of the trophectoderm (Fig. 8).

To investigate ePAD localization at the ultrastructurallevel, immunoelectron microscopy was performed by using

Fig. 8. Scanning confocal analysis of ePAD subcellular localization. Immature oocytes (GV) were collected from ovaries. For fertilized eggs (Fert.),metaphase II eggs were collected from superovulated females, the zona pellucidi were removed with chymotrypsin, and the eggs were coincubated with spermfor 45 min and fixed for evaluation. Morula and blastocysts were collected directly from females either 4 or 5 days after mating. In immature oocytes, ePADappears to be predominantly expressed in the cytoplasm; however, some nuclear staining is also evident, while no staining is seen in the nucleolus. In fertilizedeggs, ePAD is expressed in punctate patches the throughout the cytoplasm. In the morula and blastocyst (Blast.), ePAD appears to be localized mainly tothe cytoplasm; however, there does appear to be a limited amount of nuclear staining in some blastomeres. Note that the blastomere indicated by the arrowappears to be in metaphase and that ePAD is excluded from the chromatin-containing region. Also, cortical staining is seen at each of the developmental stagesshown. In the morula, however, the cortical staining is limited to the external surface of the blastomeres and appears polarized. In the blastocyst, the corticalstaining is mainly limited to the trophectoderm. IIF, indirect immunofluorescence; Ph., phase contrast.


anti-ePAD-coated 10-nm gold particles to probe ovarianand cumulus–oocyte complex cross-sections (Fig. 9). Com-parison of low magnification fields revealed that the preim-mune IgG-coated gold particles did not bind to the egg

cytoplasm, while the anti-ePAD IgG gold particles localizedwithin the egg cytoplasm to whorls of electron dense ma-terial. Anti-ePAD immunogold particles were not found tospecifically associate with any other ovarian cell type in-

Fig. 9. Immunoelectron microscopic localization of ePAD in oocytes. At low magnification (LM) using 10-mm-labeled gold particles, few preimmune-coated(PI) particles are seen in a cytoplasmic section of an egg, while immune-coated particles (I) are found to be abundantly associated with the egg cytoplasmin a swirling pattern. Higher magnification (HM) using 5-mm gold particles reveals few preimmune particles in the cytoplasmic region of the egg, whilenumerous immune particles predominantly localize to the cytoplasmic sheets of the egg. At high magnification, 10-mm particles coated with anti-ePAD IgGclearly localize to the sheets. This cytoskeletal structure is unique to the mammalian egg and embryo and is thought to be mainly composed of intermediatefilaments.


vestigated, nor did the particles appear to specifically asso-ciate with electron lucent regions of cytoplasm. At highermagnifications, anti-ePAD immunogold particles localizedto multilamellar filamentous cytoskeletal structures thathave been previously named cytoplasmic sheets. These cy-toskeletal specializations are distinctive in their lamellarappearance within the cytoplasm of the egg and early em-bryo (Capco et al., 1993).

Discussion

Our laboratory has been utilizing proteomics to charac-terize proteins in the mouse egg in an effort to identify newpotential contraceptive targets. We chose to clone and char-acterize ePAD based on the observation that it representedone of the most abundant proteins yet to be characterized inthe mouse egg. The specific localization of ePAD to oocytesin ovarian sections and its homology to a well-characterizedenzyme family that has known in vitro and in vivo sub-strates supports further development of small molecule in-hibitors to this contraceptive target.

A central question germane to the potential role of ePADin egg physiology is whether ePAD functions as an eggand/or embryo peptidylarginine deiminase, and if so, whatare its substrates? Previous experiments have shown that thecytoplasmic sheets consist of a highly cross-linked networkof Triton X-100 detergent-insoluble intermediate filamentsthat are coated with a Tween 20 detergent-soluble proteincomponent. The most abundant soluble component of thesheets is an �69-kDa protein that has not been character-ized at the molecular level (Gallicano et al., 1994a). GivenePADs localization to the cytoskeletal sheets, its solubilitycharacteristics, and its molecular weight (75 vs 69 kDa), itis tempting to speculate that ePAD is the soluble �69-kDaprotein previously described in the literature (Gallicano etal., 1994a). The molecular nature of the insoluble compo-nent has been partially characterized and shown to containcytokeratin (Gallicano et al., 1994a). Because keratin issuch a well-characterized PAD substrate in epithelial cells(Senshu et al., 1999a,b), it represents an obvious potentialsubstrate for ePAD in eggs and embryos.

In epithelial cells, deimination of specific arginine resi-dues in keratin and filaggrin is known to be involved in theconversion of a preexisting fine cytokeratin network to moredensely bundled linear filamentous structures (Ishida-Yamamoto et al., 2002). This reorganization is reminiscentof changes that have been described in the cytoskeletalsheets in the egg where the preexisting multilamellar,whorl-like pattern in unfertilized eggs (similar to that seenin Fig. 9) becomes more linearly arranged in the fertilizedzygote (Gallicano et al., 1995). At compaction, the sheetsappear as highly polarized structures with one end anchoredon the apical plasma membrane and extending downwardtoward the basolateral surface within each blastomere. Aswith ePAD protein expression, by the blastocyst stage of

development, the sheet density begins to diminish. In thetrophectoderm, the sheets are localized near the plasmalamina, while in the inner cell mass blastomeres, the sheetstend to be more centrally located (Capco et al., 1993).

Because PADs are calcium-dependent enzymes, fertili-zation represents a particularly interesting developmentalstage for potential ePAD activity since large calcium tran-sients are known to occur following sperm–egg fusion (Car-roll, 2001). Following fertilization, one could imagine thatthe calcium transients could activate ePAD, which wouldthen deiminate keratin, resulting in the observed cytoskel-etal sheet reorganization. Such reorganization may be es-sential for sperm internalization or polar body extrusion.

In addition to the keratins of the cytoskeletal sheets, thenuclear localization of ePAD suggests that other proteinsalso represent potential ePAD deimination targets at fertil-ization. For example, protamine, an arginine-rich (�60%arginine residues) sperm-specific histone-like protein, isknown to be an excellent in vitro substrate for PADs (Sug-awara et al., 1982). Given that substrate deimination causeschanges in secondary structure (Tarcsa et al., 1996), deimi-nation of sperm protamine by ePAD, following gametefusion, might facilitate nuclear decondensation. In fact, wehave found that commercially available skeletal musclePAD readily decondenses sperm chromatin in vitro (datanot shown). Further experiments are being performed todetermine whether either native or full-length recombinaintePAD can decondense sperm in vitro and in vivo.

Histone proteins in HL-60 granulocytes have recentlybeen found to contain citrulline residues and therefore arelikely substrates for PADs (Hagiwara et al., 2002). Theinvestigators who made this observation hypothesized thatthe N-terminal tails represent the most likely site for PADactivity, and if this prediction is correct, deimination of thehistone tails represents a novel posttranslational histonemodification. Other N-terminal histone modifications, suchas acetylation, phosphorylation, and methylation, have beenshown to alter chromatin structure and thus lead to changesin transcriptional activity of specific genes (Jenuwein andAllis, 2001). It is well documented that there is a dramaticreprogramming of gene expression patterns following fer-tilization and prior to zygotic gene activation; however, themechanism by which the egg is able to perform this remark-able task is unknown. The egg factors involved in thisprocess are likely to be cytoplasmic, as demonstrated by theobservation that somatic cell reprogramming requires thatthe nuclei be placed in the context of the egg cytoplasm(Surani, 2001). It will be of interest to establish whetherePAD can deiminate egg/embryo histones and thus possiblyplay a role in the reprogramming process.

To conclude, we have cloned and characterized ePAD, ahighly abundant egg and embryo protein that is �40%identical to the peptidylarginine deiminase enzyme family.ePAD appears to be expressed from the primary oocytestage of oogenesis until at least the blastocyst stage ofdevelopment. At the ultrastructural level, ePAD localizes to


the egg cytoplasmic sheets, keratin-containing structuresknown to undergo changes in their structure during earlydevelopment. As one of the most abundant proteins in theegg proteome, further consideration of the functions of thispotentially novel deiminase are germane to fertilization andearly development.

Acknowledgments

This work was supported by NIH Grants HD 38353 andU54 29099. We would also like to thank Kenneth Klotz forhis assistance.

References

Akiyama, K., Sakurai, Y., Asou, H., Senshu, T., 1999. Localization ofpeptidylarginine deiminase type II in a stage-specific immature oligo-dendrocyte from rat cerebral hemisphere. Neurosci. Lett. 274, 53–55.

Akiyama, K., Senshu, T., 1999. Dynamic aspects of protein deimination indeveloping mouse epidermis. Exp. Dermatol. 8, 177–186.

Angerer, L.M., Cox, K.H., Angerer, R.C., 1987. Demonstration of tissue-specific gene expression by in situ hybridization. Methods Enzymol.152, 649–661.

Asaga, H., Ishigami, A., 2001. Protein deimination in the rat brain afterkainate administration: citrulline-containing proteins as a novel markerof neurodegeneration. Neurosci. Lett. 299, 5–8.

Capco, D.G., Gallicano, G.I., McGaughey, R.W., Downing, K.H., Larabell,C.A., 1993. Cytoskeletal sheets of mammalian eggs and embryos: alattice-like network of intermediate filaments. Cell Motil. Cytoskeleton24, 85–99.

Capco, D.G., McGaughey, R.W., 1986. Cytoskeletal reorganization duringearly mammalian development: analysis using embedment-free sec-tions. Dev. Biol. 115, 446–458.

Carroll, J., 2001. The initiation and regulation of Ca2� signalling atfertilization in mammals. Semin. Cell Dev. Biol. 12, 37–43.

Coonrod, S.A., Bolling, L.C., Wright, P.W., Visconti, P.E., Herr, J.C.,2001. A morpholino phenocopy of the mouse MOS mutation. Genesis30, 198–200.

Coonrod, S.A., Naaby-Hansen, S., Shetty, J., Shibahara, H., Chen, M.,White, J.M., Herr, J.C., 1999. Treatment of mouse oocytes with PI-PLCreleases 70-kDa (pI 5) and 35- to 45-kDa (pI 5.5) protein clusters fromthe egg surface and inhibits sperm–oolemma binding and fusion. Dev.Biol. 207, 334–349.

Gallicano, G.I., Capco, D.G., 1995. Remodeling of the specialized inter-mediate filament network in mammalian eggs and embryos duringdevelopment: regulation by protein kinase C and protein kinase M.Curr. Top. Dev. Biol. 31, 277–320.

Gallicano, G.I., Larabell, C.A., McGaughey, R.W., Capco, D.G., 1994a.Novel cytoskeletal elements in mammalian eggs are composed of aunique arrangement of intermediate filaments. Mech. Dev. 45, 211–226.

Gallicano, G.I., McGaughey, R.W., Capco, D.G., 1991. Cytoskeleton ofthe mouse egg and embryo: reorganization of planar elements. CellMotil. Cytoskeleton 18, 143–154.

Gallicano, G.I., McGaughey, R.W., Capco, D.G., 1992. Cytoskeletal sheetsappear as universal components of mammalian eggs. J. Exp. Zool. 263,194–203.

Gallicano, G.I., McGaughey, R.W., Capco, D.G., 1994b. Ontogeny of thecytoskeleton during mammalian oogenesis. Microsc. Res. Tech. 27,134–144.

Gallicano, G.I., McGaughey, R.W., Capco, D.G., 1995. Protein kinase M,the cytosolic counterpart of protein kinase C, remodels the internalcytoskeleton of the mammalian egg during activation. Dev. Biol. 167,482–501.

Hagiwara, T., Nakashima, K., Hirano, H., Senshu, T., Yamada, M., 2002.Deimination of arginine residues in nucleophosmin/B23 and histones inHL-60 granulocytes. Biochem. Biophys. Res. Commun. 290, 979–983.

Ishida-Yamamoto, A., Senshu, T., Eady, R.A., Takahashi, H., Shimizu, H.,Akiyama, M., Iizuka, H., 2002. Sequential reorganization of cornifiedcell keratin filaments involving filaggrin-mediated compaction andkeratin 1 deimination. J. Invest. Dermatol. 118, 282–287.

Jenuwein, T., Allis, C.D., 2001. Translating the histone code. Science 293,1074–1080.

Kanno, T., Kawada, A., Yamanouchi, J., Yosida-Noro, C., Yoshiki, A.,Shiraiwa, M., Kusakabe, M., Manabe, M., Tezuka, T., Takahara, H.,2000. Human peptidylarginine deiminase type III: molecular cloningand nucleotide sequence of the cDNA, properties of the recombinantenzyme, and immunohistochemical localization in human skin. J. In-vest. Dermatol. 115, 813–823.

Laemmli, U.K., 1970. Cleavage of structural proteins during the assemblyof the head of bacteriophage T4. Nature 227, 680–685.

Lamensa, J.W., Moscarello, M.A., 1993. Deimination of human myelinbasic protein by a peptidylarginine deiminase from bovine brain.J. Neurochem. 61, 987–996.

Lehtonen, E., 1985. A monoclonal antibody against mouse oocyte cy-toskeleton recognizing cytokeratin-type filaments. J. Embryol. Exp.Morphol. 90, 197–209.

Lehtonen, E., 1987. Cytokeratins in oocytes and preimplantation embryosof the mouse. Curr. Top. Dev. Biol. 22, 153–173.

Lehtonen, E., Lehto, V.P., Vartio, T., Badley, R.A., Virtanen, I., 1983.Expression of cytokeratin polypeptides in mouse oocytes and preim-plantation embryos. Dev. Biol. 100, 158–165.

Moscarello, M.A., Pritzker, L., Mastronardi, F.G., Wood, D.D., 2002.Peptidylarginine deiminase: a candidate factor in demyelinating dis-ease. J. Neurochem. 81, 335–343.

Nagata, S., Yamagiwa, M., Inoue, K., Senshu, T., 1990. Estrogen regulatespeptidylarginine deiminase levels in a rat pituitary cell line in culture.J. Cell Physiol. 145, 333–339.

Nakashima, K., Hagiwara, T., Ishigami, A., Nagata, S., Asaga, H.,Kuramoto, M., Senshu, T., Yamada, M., 1999. Molecular character-ization of peptidylarginine deiminase in HL-60 cells induced by reti-noic acid and 1alpha,25-dihydroxyvitamin D(3). J. Biol. Chem. 274,27786–27792.

Nishijyo, T., Kawada, A., Kanno, T., Shiraiwa, M., Takahara, H., 1997.Isolation and molecular cloning of epidermal- and hair follicle-specificpeptidylarginine deiminase (type III) from rat. J. Biochem. (Tokyo)121, 868–875.

Pratt, H.P., Bolton, V.N., Gudgeon, K.A., 1983. The legacy from theoocyte and its role in controlling early development of the mouseembryo. Ciba Found. Symp. 98, 197–227.

Rasmussen, H.H., Van Damme, J., Puype, M., Gesser, B., Celis, J.E.,Vandekerckhove, J., 1991. Microsequencing of proteins recorded inhuman two-dimensional gel protein databases. Electrophoresis12, 873–882.

Reddi, P.P., Castillo, J.R., Klotz, K., Flickinger, C.J., Herr, J.C., 1994.Production in Escherichia coli, purification and immunogenicity ofacrosomal protein SP-10, a candidate contraceptive vaccine. Gene 147,189–195.

Schultz, R.M., Davis Jr., W., Stein, P., Svoboda, P., 1999. Reprogrammingof gene expression during preimplantation development. J. Exp. Zool.285, 276–282.

Senshu, T., 1990. [Recent progress in peptidylarginine deiminase re-search]. Seikagaku 62, 192–196.

Senshu, T., Akiyama, K., Ishigami, A., Nomura, K., 1999a. Studies onspecificity of peptidylarginine deiminase reactions using an immuno-chemical probe that recognizes an enzymatically deiminated partialsequence of mouse keratin K1. J. Dermatol. Sci. 21, 113–126.


Senshu, T., Akiyama, K., Nagata, S., Watanabe, K., Hikichi, K., 1989.Peptidylarginine deiminase in rat pituitary: sex difference, estrous cy-cle-related changes, and estrogen dependence. Endocrinology 124,2666–2670.

Senshu, T., Akiyama, K., Nomura, K., 1999b. Identification of citrullineresidues in the V subdomains of keratin K1 derived from thecornified layer of newborn mouse epidermis. Exp. Dermatol. 8,392– 401.

Sugawara, K., Oikawa, Y., Ouchi, T., 1982. Identification and properties ofpeptidylarginine deiminase from rabbit skeletal muscle. J. Biochem.(Tokyo) 91, 1065–1071.

Surani, M.A., 2001. Reprogramming of genome function through epige-netic inheritance. Nature 414, 122–128.

Takahara, H., Kusubata, M., Tsuchida, M., Kohsaka, T., Tagami, S.,Sugawara, K., 1992. Expression of peptidylarginine deiminase in theuterine epithelial cells of mouse is dependent on estrogen. J. Biol.Chem. 267, 520–525.

Takahara, H., Oikawa, Y., Sugawara, K., 1983. Purification and charac-terization of peptidylarginine deiminase from rabbit skeletal muscle.J. Biochem. (Tokyo) 94, 1945–1953.

Takahara, H., Okamoto, H., Sugawara, K., 1986. Affinity chromatographyof peptidylarginine deiminase from rabbit skeletal muscle on a columnof soybean trypsin inhibitor (Kunitz)-Sepharose. J. Biochem. (Tokyo)99, 1417–1424.

Tarcsa, E., Marekov, L.N., Mei, G., Melino, G., Lee, S.C., Steinert, P.M.,1996. Protein unfolding by peptidylarginine deiminase. Substrate spec-ificity and structural relationships of the natural substrates trichohyalinand filaggrin. J. Biol. Chem. 271, 30709–30716.

Uranga, J.A., Tobalina, R.G., Alonso, A., Arechaga, J., 1995. ENDO Acytokeratin expression in the inner cell mass of parthenogenetic mouseembryos. Int. J. Dev. Biol. 39, 659–662.

Yanagimachi, R., 1994. Mammalian Fertilization, in: Knobil, E., Neill, J.D.(Eds.), Physiology of Reproduction, Raven Press, New York, pp. 189–317.


ePAD, an oocyte and early embryo-abundant peptidylarginine...

Documents

Transcript of ePAD, an oocyte and early embryo-abundant peptidylarginine...